Withdrawal of Nonsteroidal Anti-Inflammatory Drugs in Rheumatoid Arthritis Patients with Low Disease Activity

Abstract

Background/Purpose: Although nonsteroidal anti-inflammatory drugs (NSAIDs) are effective in relieving joint pain in rheumatoid arthritis (RA) patients, long-term use of NSAIDs can cause adverse effects. We aimed to examine the patient-reported outcome (PRO) and its associated clinical factors in RA patients with low disease activity (LDA) after discontinuing NSAIDs.

Methods: This study was a 16-week, multi-center prospective open-label trial. RA patients who achieved LDA (28-joint Disease Activity Score [DAS28] < 3.2) who were on NSAIDs for more than a month discontinued the NSAIDs. Acetaminphen (AAP) was used as the rescue medication. Changes of DAS28 and PRO including pain visual analogue scale (pain-VAS) and Routine Assessment of Patient Index Data 3 (RAPID-3) score were assessed. NSAID was restarted when patient’s pain was intolerable with AAP. The endpoint was to analyze the group of patients who continued to withdraw NSAID. Patients were further classified to have “sustained effectiveness” who met the following: 1) pain-VAS ≤30 mm at week16 or increase less than 25% from baseline and 2) RAPID-3 score ≤6 at week16 or increase less than 25% from baseline.

Results: A total of 109 RA patients with LDA were enrolled in the study. At the end of the study, 89 (84.8%) patients had remained without restarting NSAID. In these patients, there was a difference in pain-VAS between baseline and week 16 (P= 0.010). However, changes in RAPID-3 and DAS28 were insignificant (P= 0.128 for RAPID-3 and P= 0.638 for DAS28) Moreover, 66 patients ended up to show sustained effectiveness without restarting NSAID. After adjustments of covariables, we found out joint swelling was the detrimental factor in NSAID withdrawal (odd ratio [OR] 0.150, 95% confidence interval [CI] 0.034-0.666, P= 0.013) and sustained effectiveness (OR 0.312, 95% CI 0.101-0.964, P= 0.043).

Conclusion: NSAIDs can be attempted to discontinue in RA patients with LDA, especially in patients without joint swelling.